Trend Analysis of Malaria Mortality in Ghana from 1980–2023: A Flexible Statistical Modelling Approach

Abstract

Malaria is a public health issue that has continued to be a significant challenge in Ghana despite decades of continuous control interventions. This research aimed to measure the nonlinear dynamics of age-standardised malaria mortality in Ghana over the long run and to overcome the statistical constraints of past trend analyses, which are based on linear or segmented models and do not account for uncertainty in burden estimates. Its goals were to describe sex-specific mortality patterns, determine multi-phase temporal regimes, and generate uncertainty-constrained inference on policy evaluation. Based on the standardised age-specific malaria mortality between 1980 and 2023 in Ghana, using the Global Burden of Disease, a generalised additive model with inverse-variance weighting was estimated to fit nonlinear time curves and adjust for sex disparities. The smooth temporal variation was essential, and the degrees of freedom were 11.07 and F = 270.35 (p < .001). Also, the 96.65% of the deviance was explained by the model, which had an adjusted R² of 0.963. Again, the death rates between females were also significantly lower than those of males (B = -9.31, SE = 0.99, p =.001), with both-sexes series falling between them. The highpoint of mortality was in the early 2000s, when it was close to 130-140 deaths per 100,000, and then reduced to 55-65 per 100,000 by 2023. This study presents a nonlinear modelling framework that is more robust and better than the linear trend methods employed in earlier studies. The conclusions favour long-term ITN and case management coverage, the incorporation of climate-informed early warning systems, and enhancements to mortality surveillance to accelerate the goal of malaria elimination under SDG 3.